Document 12625019

Finding and Analyzing the Location of the Ion Convection Reversal Boundary in Polar Regions using DMSP Satellite Data
Authors: Phoebe Tengdin (Mount Holyoke College), Barbara Emery, Astrid Maute, Delores Knipp, Liam Kilcommons (High Altitude Observatory)
06/1998 – 07/1999 Boundaries
1996 Boundaries
Comparison with Weimer Models
We ran the Weimer 2005 and 2001 models to compare the size of
the convection radius as measured by DMSP with the predicted
convection radius from the model given the geophysical parameters
for the DMSP data
Below: plots of the Weimer models for geophysical conditions given
by a single pass of DMSP data (overlaid in pink)
Figure 1
|Bz| strength
Figure 2
|Bz| strength
Data Analysis
Defense Meteorological Satellite Programʼs (DMSP) F13 Satellite
Sun-synchronous orbit in dusk-dawn meridian from 0600-1800 LT
Circular orbit at altitude of ~850 km
Orbital period of 101 minutes
Measurements taken by an ion drift meter
Data corresponded to 4-s averages
Data Filtering:
Northern hemisphere only
Magnetic latitudes above 45 degrees
IMF conditions for each pass were found using 5 minute OMNI data,
averaged over the pass, allowing a 10 minute lag
(to account for propagation to the ionosphere)
IMF conditions had a strong Bz southward component and low By component
Start and end of the pass must fall within 2 hours of dawn and dusk (in MLT)
Smoothing with 20 second averages
Figure 1: 07/98-06/99 Boundaries plotted against MLT, MLAT and |Bz| strength
Figure 2: 1996 Boundaries plotted against MLT, MLAT and |Bz| strength
Figures 3 - 6: Crossing radius for each pair of boundaries plotted against Bz strength or electric field (Bz*VSW)
with linear best fit line and cross correlation
Figure 3: 07/98 – 06/99 Angle vs. Bz
Figure 4: 07/98 – 06/99 Angle vs. Bz*VSW
Figure 5: 1996 Angle vs. Bz
Figure 6: 1996 Angle vs. Bz.*VSW
Statistical Comparisons between DMSP data
and the Weimer models
Weimer 2005
Figure 4
Figure 5
Figure 6
Electric Potential
Electric Potential
Measurements from the Defense Meteorological Satellite Programʼs (DMSP) F13 Satellite were used for a study of convection location in
the polar regions of the earthʼs ionosphere. The satellite was in a sun-synchronous orbit in the dawn-dusk meridian. The data available
corresponded to 4 second averages. Using a 20-second smoothing technique, the zero crossings in the ion drift velocity values were
determined for each pass of the satellite. OMNI data were used to calculate the average IMF conditions for each pass from 5 minute data
with a 10 minute lag to allow the IMF signal to propagate to the ionosphere. This study only considered northern hemisphere passes with
a Bz southward component where the By component was three times smaller than the Bz component. These passes were archived,
along with information about the location of the convection reversal boundary (zero crossing in ion drift velocity) for each pass. This
survey was conducted over two years of data (1996 and July 1998-June 1999). The data were then plotted to analyze trends in boundary
location dependent upon IMF conditions. We also ran the Weimer 2001 and 2005 models under the same geophysical conditions in
order to compare these models with our values. Preliminary results suggest that Weimer 2005 has a better correlation to the true
convection reversal boundary than Weimer 2001. Both models had higher accuracy levels on the dawn side of the earth than on the dusk
side. We also found a strong tendency in both models to place the convection reversal boundary too far towards the poles. Further
analysis will be needed to confirm these findings and complete the survey for all 15 years of DMSP F13 data and both hemispheres.
Weimer 2001
60, 8%
Figure 7
85, 12%
196, 26%
179, 26%
419, 62%
507, 66%
Good Fit
Good Fit
All “good fit” Weimer data falls within 2 degrees magnetic latitude of our DMSP crossing data.
Poor fit data is broken into crossings that fell too far equator ward and too far pole ward
2005 Dawn Side
Figure top left:
Plot for a single pass of DMSP Data
Red line – raw data
Blue line – detrended data
Green line – detrended and averaged data
Black line – zero line
Zero crossings were found and saved for each plot
Zero Crossings
Dusk side
Dawn Side
Figure bottom left:
Image showing one day of passes
of DMSP F13 on the northern hemisphere
There is a strong correlation between Bz strength and the location of the CRBs
Dawn side predictions of the Weimer models were more accurate than the dusk side (somewhat unexpected)
The Weimer models often predict the convection reversal boundary to lie too far poleward
The DMSP data has a greater range in MLAT values than the Weimer models predict
These results warrant a full scale study covering both hemispheres and all of the DMSP F13 data (1995-2009)
Further research will be necessary to verify these results
More Comparisons
48, 11%
2005 Dusk Side
23, 6%
74, 16%
122, 32%
333, 73%
242, 62%
Good Fit
Good Fit
2001 Dawn Side
2001 Dusk Side
Barbara Emery, Astrid Maute, Delores Knipp, Liam Kilcommons
High Altitude Observatory
National Center for Atmospheric Research
Marty Snow & Erin Wood, REU Coordinators
Laboratory for Atmospheric Sciences REU Program 2011
University of Colorado, Boulder
60, 16%
25, 5%
177, 37%
93, 24%
234, 60%
282, 58%
Good Fit
Good Fit
The left hand charts show that both the
Weimer models predict the dawn side
crossing much more accurately
than the dusk.
The orange and the blue pie slices
indicate whether the DMSP data fell
further equatorward or poleward
than the Weimer models predicted